A Survey of Methods and Strategies in Character Segmentation
IEEE Transactions on Pattern Analysis and Machine Intelligence
Pattern classification: a unified view of statistical and neural approaches
Pattern classification: a unified view of statistical and neural approaches
Improving Performance in Neural Networks Using a Boosting Algorithm
Advances in Neural Information Processing Systems 5, [NIPS Conference]
Fast Address Block Location on Handwritten and Machine Printed Mail--piece Images
ICDAR '97 Proceedings of the 4th International Conference on Document Analysis and Recognition
Form--Based Localization of the Destination Address Block on Complex Envelopes
ICDAR '97 Proceedings of the 4th International Conference on Document Analysis and Recognition
| Hi-index | 0.00 |
In this article, we describe the OCR and imageprocessing algorithms usedto read destination addresses from non-standard letters (flats) bySiemens postal automation system currently in use bythe Deutsche Post AG^1.We first describe the sorting machine, its OCR hardware and the sequence ofimage processing and pattern recognition algorithms needed to solvethe difficult task of reading mail addresses, especiallyhandwritten ones. The article concentrates mainly on the twoclassifiers used to recognize handprinted digits. One of them isa complex time delayed neural network(TDNN) used to classify scaled digit-features. The other classifierextracts the structure of each digit and matches it to a number of prototypes.Different digits represented by the same graph are then discriminated byclassifiying some of the features of the digit-graph with small neuralnetworks.We also describe some approaches for the segmentation of the digits in theZIP code, so that the resulting parts can be processed and evaluated by the classifiers.